Research map / Federated learning and privacy / Personalized and heterogeneous FL
FedCDWA: Decoupled Federated Prototype Distillation with Hierarchical Wasserstein Aggregation
Zhenshen Liu, Kai Fan, Wenjie Li, Kuan Zhang, HUI LI, Yintang Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use FedCDWA instead of FedProto | federated prototype learning · SVHN (10 clients) | Local accuracy: 85.92 -> 87.56 | Improved |
| use FedCDWA instead of FedAvg | federated learning · CIFAR-10 (10 clients) | Local accuracy: 50.75 -> 61.25 | Improved |
| use FedCDWA instead of FedGMKD | federated prototype distillation · CIFAR-100 (10 clients) | Global accuracy: 16.26 -> 18.01 | Improved |
| fewer clients (stronger heterogeneity) with FedCDWA | FedCDWA · CIFAR-100 | Local accuracy gain over FedGMKD: +4.11% at 10 clients vs +1.06% at 50 clients | Improved |
| FedCDWA vs FedHKD on CIFAR-10 global accuracy at 10 clients | FedCDWA · CIFAR-10 (10 clients) | Global accuracy: 49.73 -> 51.38 | Improved |
| FedCDWA vs FedProx local accuracy at 50 clients | FedCDWA · CIFAR-10 (50 clients) | Local accuracy: 67.49 -> 67.74 | Improved |
| FedCDWA adds computational overhead compared with FedAvg | FedCDWA · CIFAR-10 | Per-round time per client (s): 2.67 -> 3.35 | Hurt |
| FedCDWA vs FedAvg time on SVHN | FedCDWA · SVHN | Per-round time per client (s): 2.49 -> 2.49 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FedGMKD — FedGMKD: An Efficient Prototype Federated Learning Framework through Knowledge Distillation and Discrepancy-Aware Aggregation (NeurIPS 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Feature Distillation is the Better Choice for Model-Heterogeneous Federated Learning (NeurIPS 2025 · similarity 0.88)
- Rethinking the Temperature for Federated Heterogeneous Distillation (ICML 2025 · similarity 0.88)
- FedGMKD: An Efficient Prototype Federated Learning Framework through Knowledge Distillation and Discrepancy-Aware Aggregation (NeurIPS 2024 · similarity 0.87)
- Efficient Federated Learning via Clients-to-Server Knowledge Distillation (Student Abstract) (AAAI 2025 · similarity 0.87)
- Reviving Stale Updates: Data-Free Knowledge Distillation for Asynchronous Federated Learning (NeurIPS 2026 · similarity 0.86)
- One-shot Federated Learning via Synthetic Distiller-Distillate Communication (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
